Despite decades of intensive nervous system (CNS) neoplasms remains very poor. Median survival for adults with the most common form of CNS tumor, the cerebral glioblastoma, is 8-12 months after diagnosis. Occasional responses to single or multiple agent chemotherapy are seen in the setting of recurrent tumor, but these responses are generally of short duration, and cures are rare. Identification of agents active against glial malignancies is challenging, with no drug tested to date reliably producing responses in a majority of treated patients. Gene amplification, related to increasing grade of glioma malignancy, has been found to occur in approximately 50 percent of all glioblastoma multiforme (GBM) cases. Although amplification of N-myc and gli (2-4 percent overall) has been reported by different groups, amplification of these genes and c-myc or K-ras are considered sporadic as compared to the amplification of c-erb 1, or the epidermal growth factor (EGFR) gene. The EGFR gene, 110 kb in size, 26 exons in organization, is localized to chromosome arm 7pll-13. Beginning with the initial description of EGFR gene amplification by Libermann et al (1985), subsequent studies have confirmed that approximately 37-58 percent of GBMs, but only isolated anaplastic astrocytomas, amplify the EGFR gene. ZD 1839 is a potent inhibitor in vitro of EGFR tyrosine kinase, competitive with ATP, and noncompetitive with peptide substrate. ZD 1839 inhibits the proliferation of EGF-stimulated KB oral squamous carcinoma cells. This effect is readily reversible on removal of the compound. Enzyme inhibition appears to be selective, with little activity against other kinases tested. Growth inhibition in vivo of a wide variety of human tumour xenograft models in nude mice was demonstrated at a range of once daily, oral doses between 12.5 and 200 mg/kg per day for up to 4 months. In some already established tumours treatment with ZD 1839 produced significant regressions. From the xenograft studies, it is not yet clear if there is a correlation between the level of EGFR expression and antitumor response. The specific aims of this proposal are: 1) To identify the activity and toxicity of ZD 1839 in the treatment of adults with glioblastoma multiforme in first relapse; 2) to determine if qualitative and quantitative levels of genotypic and phenotypic EGFR expression predict response of GBM to ZD 1839.